1. Introduction
According to the regional disaster assessment framework, flash flood risk refers to the potential damage caused by flood disaster occurrences in mountainous, hilly, and river valley areas due to heavy precipitation (Shi 2016; Zhang et al. 2006); obviously, flash flood floods represent the major form of flood disasters. Changes in regional thermodynamic processes caused by climate warming have resulted in an increase in atmospheric water vapor content, and the annual precipitation and precipitation intensity is expected to significantly increase in China by the end of the twenty-first century (Zhao et al. 2014; Chen 2013; Zhu et al. 2020). Flash flood disaster events are closely related to climate change and occur frequently. From 1978 to 2014, China’s urbanization rate increased from 17.9% to 52.6% and has continued to rise, implying that economic property affected by natural disasters may increase substantially (Bai et al. 2014; Liu et al. 2015; Fang et al. 2017). Typhoon Fitow delivered heavy rainfall to Zhejiang Province in 2013, causing several landslides and geological catastrophes, with direct economic damage of CNY 4.7 billion (2013 prices) in Zhejiang during the typhoon’s effect period (Xinhua News Agency 2013). In 2016, Supertyphoon Meranti brought rainfall to Zhejiang Province, causing flash floods and landslides in Wenzhou, Jinhua, Taizhou, and other river basins, resulting in CNY 3.31 billion (2016 prices) in direct economic damage (CNR news 2016). Southeastern China is densely populated, economically developed, and an important hinterland for development in China; one of the main challenges the region must address in the context of climate change is the improvement in regional adaptation and flood disaster risk reduction (Hallegatte et al. 2013; Liu et al. 2015). The risk assessment of flash floods can inform the application of measures to enhance regional adaptive capacity.
The assessment of flash flood risk is based on the analysis of hazard intensity and its spatial distribution characteristics (Pradhan 2010; Liu et al. 2018; Wang et al. 2021). Hydrological and hydrodynamic models can effectively combine regional meteorological elements with subsurface characteristics and are frequently used in flash flood hazard assessments. The common numerical simulation models include Hydrologic Engineering Center’s Hydrologic Modeling System (HEC-HMS), HEC River Analysis System (HEC-RAS), 2D Flood Routing Model (FLO-2D), SWAT, and the Catchment-Based Macroscale Floodplain (CaMa-Flood) model (Gyawali and Watkins 2013; Peng and Lu 2013; Murty et al. 2014; Zhang et al. 2021). Combined with the simulation results of the hydrological–hydraulic model, the impact of land-use change on flash floods can be reflected. Mawasha and Britz (2020) used the SWAT hydrological model to simulate the inundation characteristics of the Jukskei River catchment based on the changes in surface characteristics. Their results revealed that the depth of surface runoff is synchronized with the increase in constructed land area. Hydrological–hydraulic models can also be combined with climate models to reflect the hazard change in flash floods. Zhang et al. (2019) simulated the inundation characteristics of flash floods in the river basin of northern China by coupling HEC-HMS and FLO-2D and showed that the increase in inundation depth over 3 m is 7.8% and 22.0% for the RCP4.5 and RCP8.5 scenarios, respectively. The specificity of flash flood disaster areas prevents researchers from understanding the inundation process and characteristics, and hydrological–hydraulic models can provide technical methodological support for hazard assessment in flash-flood-prone areas (Peng and Lu 2013; Chen et al. 2022).
Using the framework of regional disaster assessment, flash flood risk can be expressed quantitatively through economic damage and population damage; related studies mainly focus on economic damage assessment (Armal et al. 2020). Since flash floods are a form of flood disaster, their economic damage assessment methods are basically the same as flood disasters. Housing replacement cost (Ciurean et al. 2017), land-use value (Chen et al. 2020), and GDP (Mokrech et al. 2015) are frequently used to describe the characteristics of regional exposure in risk assessments. These indicators refer to the amount of cumulative change in production activity in a region over a certain period, and while they have significant time dimension characteristics, they are insufficient to express the value characteristics of regional exposure (Wu et al. 2014). In the global assessment report of disaster risk reduction (2013), the United Nation Office for Disaster Risk Reduction (UNDRR) recommends using asset value as the regional exposure indicator to maximize the depiction of economic damage caused by natural disasters. Asset value refers to the amount of capital accumulated in a region during a certain period and includes physical assets such as buildings, infrastructure, and natural resources (Wu et al. 2018; Gunasekera et al. 2015). Wu et al. (2016), based on the regional asset value, used the questionnaire method to conduct research on the flood damage rate at the town level in southeastern China. Their findings revealed a significant relationship between the disaster damage of five types of industrial areas (significance level p < 0.05). Therefore, the analysis of flash flood risk assessment based on asset value and industrial structure has scientific and practical significance (Henriet et al. 2012).
In this study, we attempted to quantitatively carry out a risk assessment of flash flood disasters within the context of economic change scenarios. We selected the Yantanxi River basin, southeastern China, as the case area, where two main objectives need to be addressed. The first objective is to obtain the risk characteristic of a certain flash flood disaster event. The second objective is to obtain the flash flood risk change within the economic scenarios. To achieve our research objectives, first, we coupled the HEC-HMS and FLO-2D models and combined them with fieldwork to form the regional flash flood hazard assessment model. Then, the asset value was spatialized with the help of industrial structure and land-use data, and we predicted the regional economic characteristic changes based on the localized shared socioeconomic pathways (SSPs) dataset. We also modeled the response of different industries to flash floods and performed a quantitative flash flood risk assessment. The results of this paper can provide new information for the planning and construction of regional flood control facilities.
2. Data and methods
a. Study area
The Yantanxi River basin (YRB), a second-level river basin of the Nanxi River, is located in Yongjia County, Zhejiang Province, southeastern China, which is a second-level river basin of the Nanxi River. The YRB area is 687.62 km2, with a typical mountainous hilly terrain (Fig. 1). The Shizhu hydrological station records the hydrological characteristics upstream of the Nanxi River upstream, including the Yantanxi River. The average annual precipitation in the Nanxi River basin is 1704.8 mm, and the average annual temperature ranges from 14.4° to 18.2°C (Dai 2006). The YRB rainy season is influenced by the passage of the rainy zone and local circulation, with May–June being the primary rainy season and typhoon rains occurring from July to September. The YRB covers four administrative units in Yantan town and is well known for its dense distribution of ancient villages and the Sihai Mountain Forest Park (Chen et al. 2022). The abundant precipitation and special topography may lead to frequent flash floods in the YRB.
Geographical location and elevation of the study area.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
Supertyphoon Lekima made landfall in China in August 2019, resulting in direct economic losses of CNY 4.24 billion in Yongjia County and 32 casualties in the YRB due to tropical cyclone–extreme precipitation–flash flood chain effects (China News 2019), thus raising concerns about the characteristics of flash floods within the region. According to meteorological station data, which include rainfall contributions to the YRB, the 2019 flash flood rainfall began at 0700 LT 9 August and ended at 1600 LT 10 August. The rainfall of the YRB during the disaster was 163.61 mm after the weight assignment calculations. The Gumbel extreme value distribution model was used to calculate the return period of this disaster event, which is 150 years in the YRB. Chen et al. (2022) pointed out that the change in flash flood impact areas is not significant in the context of climate change, while flash flood risk increases within the river basin. The authors combined geospatial statistical methods and indicated that socioeconomic factors are the main driving factor for flash flood risk in southeastern China, such as the YRB; thus, it is necessary to carry out flash flood risk assessment under economic change scenarios. Developing flash flood risk assessment in the YRB has implications for the prevention and management of similar areas prone to flash floods.
b. Data
1) Hourly precipitation data
Observed rainfall data were derived from the hourly precipitation dataset provided by the China Meteorological Administration (CMA). There are six meteorological stations contributing to the rainfall in the study area, namely, Jinyun, Yueqing, Linhai, Qingtian, Yongjia, and Xianju; the rainfall time series length is 1971–2019 for each station. We constructed Thiessen polygons based on the six meteorological stations (Fig. 1), used the ratio of river basin area to Thiessen polygon area as the contribution weight of rainfall, and obtained the rainfall characteristics of the study area by summing up the assigned weights.
2) Geographic information data
The geographic information data included the digital elevation model (DEM), land use, and soil type (Fig. 2). The DEM was obtained from the Geospatial Data Cloud constructed by the Chinese Academy of Sciences (http://www.gscloud.cn), which has a spatial resolution of 30 m. The land-use data were based on the Landsat-8 interpretation of the land surface characteristics of China in 2018, with a spatial resolution of 30 m, and provided by the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (http://www.resdc.cn). Soil-type distribution data were obtained from the Harmonized World Soil Database (HWSD), which is published by the Food and Agriculture Organization (FAO) and available at the National Cryosphere Desert Data Center of China (http://www.ncdc.ac.cn).
The characteristics of (a) land use and (b) soil type in the Nanxi River basin.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
3) Socioeconomic data
The GDP statistics were derived from the Wenzhou Statistical Yearbook; the time series is 1978–2020 (Wenzhou Municipal Bureau of Statistics 2021).
The asset value could be estimated for 344 prefecture-level cities in China from 1978 to 2012 by using the perpetual inventory method (Wu et al. 2014). Based on this dataset, we developed a regional exposure analysis.
The localized SSP dataset was provided by Nanjing University of Information Science and Technology (NUIST), because it is closer to Chinese socioeconomic features (https://geography.nuist.edu.cn/2022/0420/c1971a193806/page.htm). The research team selected 2010 as the base year and collected key indicators such as fertility, mortality, current population migration rate, capital stock, labor force participation rate, and total factor productivity in China. Based on the localization parameter system, the localized SSP dataset was created. This dataset includes the GDP of China under SSP1 (sustainable path), SSP2 (middle path), SSP3 (regional competition path), SSP4 (unbalanced path), and SSP5 (fossil-fuel-based development path) scenarios, with a spatial resolution of 0.5° × 0.5°. The corrected dataset can be used to support studies on climate change impacts at regional and river basin scales (Jiang et al. 2018; Huang et al. 2019).
c. Methods
1) Flash flood hazard assessment
To derive the inundation characteristics of flash floods and form the flash flood hazard assessment, we coupled the HEC-HMS hydrologic model with the FLO-2D hydrodynamic model. Furthermore, we referred to the case area surface features (DEM, land use, and soil type) and fieldwork to establish the model parameters.
The U.S. Army Corps of Engineers developed the HEC-HMS, which is a rainfall–runoff simulation system. The HEC-HMS can be used to simulate runoff processes in river basins with limited hydrological data, such as the YRB (Zhang et al. 2019; Chen et al. 2020, 2022). The HEC-HMS divides areas into subbasins based on the DEM; calculates river runoff by combining land use, soil type, and meteorological elements (rainfall) within the river basin; and outputs clear water process lines.
The FLO-2D model is a two-dimensional dynamic model that combines flood rheological characteristics with continuity and motion equations (O’Brien et al. 1993). FLO-2D offers technical methodological assistance for the simulation of flash flooding characteristics (Peng and Lu 2013; Mishra et al. 2018). The following is the FLO-2D calculation equation, comprising
-
continuity equation
-
motion equation
where h is the flood depth, υx and υy are the flow velocities in the x and y directions, t is the flood evolution time, I is the intensity of rainfall, Sfx and Sfy are the friction slopes in the x and y directions, and Sax and Say are the riverbed slopes in the x and y directions.
The process of FFHA model construction.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
2) Asset value spatialization
The above computation demonstrates that GDP is an important indicator for identifying regional exposure characteristics; thus, the simulation effect of GDP must be examined. In the five SSP scenarios, there was a significant relationship between statistical GDP and simulated GDP (p < 0.01), with simulated GDP greater than statistical GDP from 2010 to 2019; the average relative error was 58.55% (Table 1). This result implies that, even with the existing localized SSP dataset, the error caused by spatial resolution should not be avoided. Therefore, to better reflect the economic characteristics of Yongjia County, we used the coefficient correction method to correct the localized SSP dataset. The correction coefficient was calculated based on the average ratios between simulated and statistical GDP, allowing the two errors to exist only within the study area (Zhao et al. 2017). Based on the 2010–19 GDP, the average correction coefficients for Yongjia County in the five scenarios are 0.41 and 0.42. The revised results maintain the growth rates under the different scenarios and the economic features of Yongjia County, with an average relative error change of 3.72% (Fig. 4). Using the above correction results, we applied the correction coefficients to the simulated GDP for future years.
Comparison of simulated and statistical GDP in Yongjia County (billion CNY).
Comparison of the GDP corrected results under the different scenarios.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
The main concepts for spatializing asset value in Yongjia County are as follows: The county-level asset value is divided by industry structure, the density of each industrial asset is calculated by the level-I land-use area, and the value of industrial assets in each raster is calculated using the level-II land-use area (Fig. 5). In the Chinese land-use classification system, there are 6 level-I land-use categories and 25 level-II land-use categories (Liu et al. 2010). The level-I land-use types are based on the land resources and their use attributes, including cultivated land, forestland, grassland, water area, construction land, and unused land. The level-II land-use types are classified based on the natural characteristics of land resources; for example, cultivated land is divided into paddy fields and dry land, and construction land is divided into residential, industrial and commercial land. The spatial distribution of asset value in future scenarios can also be obtained through the relationship between GDP and asset value (formula 2).
Spatializing the asset value at the county level.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
3) Flash flood damage model
Economic damage rates of flash flood disasters in the YRB (%).
3. Results
a. The hazard assessment of flash floods
We used the FFHA model to obtain the features of flash flood inundation in 2019. The area impacted by the 2019 flash floods in the YRB was 34.68 km2, accounting for 5.04% of the total river basin area. The areas with flooding depths mainly less than 2.00 m accounted for 66.61% of the affected area, and the areas with flooding depths greater than 2.00 m accounted for 33.39% (Fig. 6). To validate the accuracy of the hazard assessment result, we chose verification points from different hazard levels and conducted field research in the YRB from 27 to 30 December 2020 (Fig. 7).
The hazard assessment result of the 2019 flash flood disaster in the YRB.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
Verification points in the YRB: (a) landslides that occurred during the 2019 flash flood disaster event, (b) houses that were impacted by the flash flood, and (c) stones carried down by the flash flood.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
A comparison of the simulated and actual inundation results showed that the simulation error value ranged between 0.03 and 1.41 m. The correlation coefficient squared R2 of the simulated and actual inundation depths was 0.48 (p < 0.05), and the root-mean-square error (RMSE) was 0.74 (Table 3). We believe that the regionalized FFHA model, as compared with related research (Xu et al. 2020), reflects the inundation characteristics of the 2019 YRB flash flood event to a certain extent. Therefore, we used the simulated results to develop flash flood risk assessments in the YRB under the economic change scenarios.
Comparison of inundation depth in the 2019 flash flood disaster event (m).
b. Regional exposure characteristics
During the 2019 flash flood disaster, the asset value of Yongjia County was CNY 77.11 billion, according to the 2019 GDP. The primary, secondary, and tertiary industrial structures in Yongjia County were 3.6%, 43.1%, and 53.3%, with asset values of CNY 2.78 billion, CNY 33.23 billion, and CNY 41.10 billion, respectively.
To spatialize the asset value in Yongjia County, we combined the preceding spatialization methods; the results are shown in Fig. 8. According to the 2018 administrative division of Yongjia County, the region is divided into 17 administrative units, including Nancheng, Dongcheng, Huangtian, Oubei, Shatou, and Yantan. The asset value of Yongjia County is primarily concentrated in the region’s southwestern part, which is connected to the main urban area of Wenzhou City. The central part of Yongjia County has a few scattered concentrations of assets, while the northern mountainous areas are not significantly concentrated. We combined the northern portion of Yongjia County with field research and discovered that it has a large number of ancient villages, natural forest parks, and other tourism resources, while the tourism service facilities utilized to receive tourists are concentrated in Nancheng. Based on the spatial statistical analysis, the total assets of Yantan town account for 2.77% of the total assets in Yongjia County, which is the main administrative area contained in the YRB.
Spatialization characteristics of asset value in Yongjia County.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
The GDP and asset value characteristics of Yongjia County under future scenarios were analyzed using a time interval of 10 years. Within the five SSP scenarios, the GDP and asset value show a significant growth trend (Fig. 9). In the SSP1, SSP2, SSP3, SSP4, and SSP5 scenarios, the GDP and asset value initially experience rapid growth, and then the growth rate flattens after 2050. Until 2100, the SSP1 scenario has the highest GDP and asset value exposure, approximately CNY 220.39 billion and CNY 370.26 billion, respectively, followed by SSP3 (CNY 213.77 billion and CNY 359.13 billion) and SSP4 (CNY 179.73 billion and CNY 301.95 billion).
The characteristics of (a) GDP and (b) asset value in Yongjia County.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
We chose SSP2 (middle path) to represent the current level of economic development and SSP5 (fossil fuel-based development path) to represent the extreme level of economic development and carried out further analysis of the exposure change in Yongjia County. Under the SSP2 and SSP5 scenarios, Yongjia County’s GDP and asset value show a significant upward trend, whereas the annual growth rate shows a downward trend. Under the SSP2 and SSP5 scenarios, the GDP growth rate shows an upward trend from 2070 to 2080, and after 2080, the GDP tends to increase. The annual growth rate of GDP from 2060 to 2070 shows an upward trend in the SSP2 scenario and a downward trend in the SSP5 scenario. According to GDP statistics, the annual growth rate of GDP in Yongjia County from 2009 to 2019 was 9.71%, while from 2090 to 2100, it was 1.06% (SSP2 scenario) and 1.54% (SSP5 scenario), a decrease of 8.65% and 8.17%, respectively. In 2100, for exposure to flash floods, Yongjia County is expected to have CNY 308.32 billion in GDP and CNY 517.98 billion in asset value under the SSP2 scenario, and CNY 476.31 billion in GDP and CNY 800.21 billion in asset value under the SSP5 scenario.
c. Flash flood risk assessment
Calculated by the flash flood economic damage model, the 2019 flash flood disaster event caused CNY 0.55 billion in economic damage in the YRB, with the primary, secondary, and tertiary industries accounting for 37.19%, 32.65%, and 30.16% of the total economic damage, respectively. A total of 83.67% of the economic damage from the 2019 flash flood occurred in areas with inundation depths of up to 2.00 m. We developed an assumption of constant vulnerability of regional exposure, based on a combination of the spatial distribution characteristics of asset value from 2020 to 2100; the flash flood risk characteristics can be examined within the five economic change scenarios (Fig. 10). The risk of flash floods in the YRB increases significantly and presents exponential growth in the different SSP scenarios. In 2100, the economic damage caused by flash floods changes to CNY 4.51 billion, CNY 4.40 billion, and CNY 3.70 billion within the context of the SSP1, SSP3, and SSP4 scenarios, respectively. At the end of the twenty-first century, the economic damage caused by flash floods is CNY 6.35 billion and CNY 9.80 billion under SSP2 (current development level) and SSP5 (extreme development level), respectively, and the flash flood risk is predicted to increase by 91.34% and 94.39% relative to 2019.
The risk changes of flash flood disasters under the five SSP scenarios.
Citation: Weather, Climate, and Society 15, 1; 10.1175/WCAS-D-21-0141.1
4. Discussion
In this paper, we discussed the impact of economic factors on flash flood risk change in the YRB, southeastern China. First, we built a regional flash flood hazard assessment model. Second, we analyzed the spatial distribution characteristics of asset value using the top-down approach. Finally, we considered the responses of three industries to flash floods and used a combined downscaled SSP dataset to perform a quantitative assessment of flash flood risk under economic change scenarios.
a. Extreme characteristics of flash floods in the context of economic change scenarios
The regionalized hazard assessment model better reflects the inundation characteristics of flash floods within the YRB; the area affected by the 2019 flash flood disasters accounted for 5.04% of the total river basin area and was primarily concentrated in areas with an inundation depth of less than 2.00 m. The specificity of the region where flash floods occur and the method of coupling the HEC-HMS with the FLO-2D model can provide technical methodological support for flash flood hazard assessment in similar regions (Zhang et al. 2019; Peng and Lu 2013). In this paper, we developed a regional FFHA model based on topography, surface features, and soil types. In addition, we verified the hazard assessment results by fieldwork methods, and the regional FFHA model was able to produce better results when compared with related studies (Xu et al. 2020). The flash flood hazard assessment method we used is helpful in developing risk assessments in flash-flood-prone areas such as the YRB. However, flash floods have a distribution area of 4.87 × 106 km2 in China (Cui and Zou 2016), and socioeconomic-related assets are exposed to the constant threat of flash floods. Given the significant impact of flash floods on China and the widespread distribution of flash flood disasters, how to implement flash flood risk assessment on a large scale requires further discussion.
Under the economic change scenarios (2020–100), the asset value of exposure to flash floods in the YRB increases gradually within the five SSP scenarios. Economic drive changes in flash flood risk within the YRB and significantly increase in the five SSP scenarios, with increases of 91.34% (SSP2—current development level) and 94.39% (SSP5—extreme development level) at the end of the twenty-first century relative to 2019. In this paper, asset value is described as an exposure feature and spatialized using a top-down approach that combines land-use data with regional industrial structure (Wu et al. 2016). As a mixed economic indicator, asset value includes the characteristics of buildings, infrastructure, and natural resources that can better describe the damage caused by a natural disaster (UNDRR 2013). Based on the characteristics of asset value, we characterize the maximum impact of flash floods triggered by Lekima in the YRB. GDP is an important indicator for determining exposure characteristics, and the localized SSP dataset effect is analyzed for flash flood risk assessment. The localized SSP dataset can be used to describe the Chinese economic factors within the river basin scale (Jiang et al. 2018; Huang et al. 2019); inevitably, there are still differences in spatial resolution for the case area. Therefore, we make a secondary correction to this dataset based on the correction coefficient method to control the error within Yongjia County (Zhao et al. 2017).
b. Limitations and future research
1) Insufficient analysis of hazard assessment in future scenarios
In this paper, we conducted a hazard assessment based on the FFHA model. Although we verified the inundation result, we were unable to obtain the actual hydrological flow processes for validating the simulation ability of the HEC-HMS; this failure could introduce errors into the risk assessment results (Shamsudin et al. 2011). Meanwhile, we ignore the surface features change within the future scenarios, which also introduce error for changes in river conditions during hazard simulated results. On the other hand, the driving factor analysis was based on single climate model research (Chen et al. 2022); thus, some of the flash flood risk change characteristics were neglected. In future research, climate model ensemble methods could be used to detect more climate change features, and combined with socioeconomic changes, more extreme flash flood characteristics in future scenarios could be obtained.
2) The use of single data leads to inadequate analysis of regional exposure characteristics
In the analysis of the spatial exposure characteristics, we formed a spatialization based on single land-use data; we combined the asset value of the secondary and tertiary industries. In subsequent studies, we could add more human activity factors for the exposure analysis, such as lighting, road density, population density, and buildings (Ma et al. 2014; Gunasekera et al. 2015; Wu et al. 2018; Zheng et al. 2020). With the further subdivision of industry types, we could better obtain the internal characteristics of regional exposure (Henriet et al. 2012). Under future scenarios, the predicted exposure value was calculated using the relationship between GDP and the asset value; thus, the results did not account for the characteristics of the asset value change within the space (Ye et al. 2019). In addition, the estimated future asset value ignores the trend of the asset value itself, and land-use change in the future years within the region will also introduce uncertainty to the spatial distribution of assets. In this paper, we focused on the economic factors impacting the risk of flash floods; however, there were 32 casualties in the 2019 flash flood disaster within the YRB, and we lacked an analysis of the impact on population. Whether a larger population will be exposed to the risk of flash floods driven by the two-child policy in China is a direction we need to discuss in future research.
3) Improvement in economic damage assessment methods for flash flood disasters
Using the regional exposure spatialization method, we introduced the response of different industries to flash floods for damage assessment (Liu 2011; Wu et al. 2016; Ning et al. 2020). Note that the results of the economic damage assessment were not compared with the actual disaster situation. The inundation depth of flash floods is one of the characteristics; thus, we used the inundation depth to discuss the relationship between flash floods and economic factors. In the fieldwork, we found that the economic damage within the YRB was mainly caused by the impact of the sediment mixture brought by flash floods, which is related to flow velocity and impact force. To better reflect the characteristics of flash floods, in later studies, we can construct flash flood vulnerability curves based on the flow velocity and water impact force (Totschnig et al. 2011; Zhang et al. 2016). For the flash flood risk assessment of future scenarios, we assumed that the response of each industry to flash floods would be constant. The response of different industries to flash flood changes also needs to be considered in subsequent studies. There are many factors of uncertainty included in the risk assessment process, for example, the simulated flow process, spatialization calculation method, or disaster sample size used. It is necessary to define the range of uncertainty and describe the natural disaster risk in the form of interval thresholds, which is what we need to carry out in our subsequent research.
5. Conclusions
In this paper, we combine numerical model simulations, fieldwork, and spatial analysis methods to develop a flash flood risk assessment in the Yantanxi River basin (YRB), southeastern China, within the context of economic change. The results show that the return period of the 2019 flash flood disaster in the YRB is 150 years; the flash-flood-affected area in 2019 accounted for 5.04% of the total river basin area and was concentrated at inundation depths of less than 2.00 m. By the end of the twenty-first century, the risk of flash flooding increases by 91.34% (SSP2—current economic development level) and 94.39% (SSP5—extreme economic development level).
Based on the fact that economic factors are the main drivers of flash flood risk change in the Yantanxi River basin, we first include asset value in conjunction with industrial structure in flash flood risk assessment. The findings of our study estimate the economic damage caused by the 2019 flash flood disaster in the Yantanxi River basin; we also predict the future risk changes of an extreme disaster. Under the economic change scenarios, the risk of flash floods significantly increases, and for the high-level hazard area, management strategies for potentially affected populations and designs for shelter areas are needed. During the process of our research, the dataset in mountainous and hilly parts of southeastern China was found to be poorly constructed, with only county-level-scale exposure available in the published statistical dataset. Disaster datasets guarantee the accuracy of risk assessment results. How to strengthen the construction of a disaster database so that it can better assist with improving disaster prevention and mitigation capacity is an important challenge for future regional development.
Acknowledgments.
We thank the editors and reviewers for their comments. This work was supported by the National Key Research and Development Program (2017YFA0604903; 2017YFC1502505).
Data availability statement.
The asset value data used in this paper are available in the published research (https://doi.org/10.1016/j.chieco.2014.10.008). China localized SSPs datasets are available through Nanjing University of Information Science and Technology (https://geography.nuist.edu.cn/2022/0420/c1971a193806/page.htm).
REFERENCES
Armal, S., J. R. Porter, B. Lingle, Z. Chu, M. L. Marston, and O. E. J. Wing, 2020: Assessing property level economic impacts of climate in the US, new insights and evidence from a comprehensive flood risk assessment tool. Climate, 8, 116, https://doi.org/10.3390/cli8100116.
Bai, X., P. Shi, and Y. Liu, 2014: Society: Realizing China’s urban dream. Nature, 509, 158–160, https://doi.org/10.1038/509158a.
Chen, H., 2013: Projected change in extreme rainfall events in China by the end of the 21st century using CMIP5 models. Chin. Sci. Bull., 58, 1462–1472, https://doi.org/10.1007/s11434-012-5612-2.
Chen, L., Z. Yan, Q. Li, and Y. Xu, 2022: Flash flood risk assessment and driving factors: A case study of the Yantanxi River Basin, southeastern China. Int. J. Disaster Risk Sci., 13, 291–304, https://doi.org/10.1007/s13753-022-00408-3.
Chen, Y., Y. Wang, Y. Zhang, Q. Luan, and X. Chen, 2020: Flash floods, land-use change, and risk dynamics in mountainous tourist areas: A case study of the Yesanpo scenic area, Beijing, China. Int. J. Disaster Risk Reduct., 50, 101873, https://doi.org/10.1016/j.ijdrr.2020.101873.
China News, 2019: Typhoon “Likima” killed 23 people in Shao Village, Yongjia Mountain, Zhejiang, and 9 are missing. Accessed 20 June 2020, https://baijiahao.baidu.com/s?id=1641567986991915667&wfr=spider&for=pc.
Ciurean, R. L., H. Hussin, C. J. Van Westen, M. Jaboyedoff, P. Nicolet, L. Chen, S. Frigerio, and T. Glade, 2017: Multi-scale debris flow vulnerability assessment and direct loss estimation of buildings in the eastern Italian alps. Nat. Hazards, 85, 929–957, https://doi.org/10.1007/s11069-016-2612-6.
CNR news, 2016: Typhoon Meranti affected more than 1.5 million people in Zhejiang Province, 10 people were casualties and 4 people were missing. Accessed 20 June 2020, http://news.cnr.cn/native/gd/20160917/t20160917_523140316.shtml.
Cui, P., and Q. Zou, 2016: Theory and method of risk assessment and risk management of debris flows and flash floods (in Chinese). Prog. Geogr., 35, 137–147, https://doi.org/10.18306/dlkxjz.2016.02.001.
Dai, G., 2006: Natural disaster. Records of Water Conservancy in Yongjia (in Chinese). Zhonghua Book Company Press, 31–89.
Fang, J., W. Liu, S. Yang, S. Brown, R. J. Nicholls, J. Hinkel, X. Shi, and P. Shi, 2017: Spatial-temporal changes of coastal and marine disasters risks and impacts in mainland China. Ocean Coastal Manage., 139, 125–140, https://doi.org/10.1016/j.ocecoaman.2017.02.003.
Gunasekera, R., O. Ishizawa, C. Aubrecht, B. Blankespoor, S. Murray, A. Pomonis, and J. Daniell, 2015: Developing an adaptive global exposure model to support the generation of country disaster risk profiles. Earth-Sci. Rev., 150, 594–608, https://doi.org/10.1016/j.earscirev.2015.08.012.
Gyawali, R., and D. W. Watkins, 2013: Continuous hydrologic modeling of snow-affected watersheds in the Great Lakes Basin using HEC-HMS. J. Hydrol. Eng., 18, 29–39, https://doi.org/10.1061/(ASCE)HE.1943-5584.0000591.
Hallegatte, S., C. Green, R. J. Nicholls, and J. Corfee Morlot, 2013: Future flood losses in major coastal cities. Nat. Climate Change, 3, 802–806, https://doi.org/10.1038/nclimate1979.
Henriet, F., S. Hallegatte, and L. Tabourier, 2012: Firm-network characteristics and economic robustness to natural disasters. J. Econ. Dyn. Control, 36, 150–167, https://doi.org/10.1016/j.jedc.2011.10.001.
Huang, J., and Coauthors, 2019: Effect of fertility policy changes on the population structure and economy of China: From the perspective of the shared socioeconomic pathways. Earth’s Future, 7, 250–265, https://doi.org/10.1029/2018EF000964.
Jiang, T., J. Zhao, L.-G. Cao, Y.-J. Wang, B.-D. Su, C. Jing, R. Wang, and C. Gao, 2018: Projection of national and provincial economy under the shared socioeconomic pathways in China. Adv. Climate Change Res., 14, 50–58, https://doi.org/10.12006/j.issn.1673-1719.2017.161.
Liu, J., and Coauthors, 2010: Spatial patterns and driving forces of land use change in China during the early 21st century. J. Geogr. Sci., 20, 483–494, https://doi.org/10.1007/s11442-010-0483-4.
Liu, J., J. Wen, Y. Huang, M. Shi, Q. Meng, J. Ding, and H. Xu, 2015: Human settlement and regional development in the context of climate change: A spatial analysis of low elevation coastal zones in China. Mitigation Adapt. Strategies Global Change, 20, 527–546, https://doi.org/10.1007/s11027-013-9506-7.
Liu, Y., 2011: Multi-scale natural disaster scenarios’ risk assessment and zoning—A case study of Wenzhou city, Zhejiang Province, Shanghai. Ph.D. dissertation, East China Normal University, 185 pp.
Liu, Y., Z. Yang, Y. Huang, and C. Liu, 2018: Spatiotemporal evolution and driving factors of China’s flash flood disasters since 1949. Sci. China Earth Sci., 61, 1804–1817, https://doi.org/10.1007/s11430-017-9238-7.
Ma, T., C. Zhou, T. Pei, S. Haynie, and J. Fan, 2014: Responses of Suomi-NPP VIIRS-derived nighttime lights to socioeconomic activity in China’s cities. Remote Sens. Lett., 5, 165–174, https://doi.org/10.1080/2150704X.2014.890758.
Mawasha, T. S., and W. Britz, 2020: Simulating change in surface runoff depth due to LULC change using soil and water and assessment tool for flash flood prediction. S. Afr. J. Geomatics, 9, 282–301, https://doi.org/10.4314/sajg.v9i2.19.
Mishra, B. K., A. R. Emam, Y. Masago, P. Kumar, R. K. Regmi, and K. Fukushi, 2018: Assessment of future flood inundations under climate and land use change scenarios in the Ciliwung River basin, Jakarta. J. Flood Risk Manag., 11, S1105–S1115, https://doi.org/10.1111/jfr3.12311.
Mokrech, M., A. S. Kebede, R. J. Nicholls, F. Wimmer, and L. Feyen, 2015: An integrated approach for assessing flood impacts due to future climate and socio-economic conditions and the scope of adaptation in Europe. Climatic Change, 128, 245–260, https://doi.org/10.1007/s10584-014-1298-6.
Murty, P. S., A. Pandey, and S. Suryavanshi, 2014: Application of semi-distributed hydrological model for basin level water balance of the Ken basin of central India. Hydrol. Processes, 28, 4119–4129, https://doi.org/10.1002/hyp.9950.
Ning, S., J. Huang, Z. Wang, and H. Wang, 2020: Indirect economic losses of flood disaster based on an input-output model: A case study of Hubei Province (in Chinese). Prog. Geogr., 39, 420–432, https://doi.org/10.18306/dlkxjz.2020.03.007.
O’Brien, J. S., P. Y. Julien, and W. T. Fullerton, 1993: Two-dimensional water flood and mudflow simulation. J. Hydraul. Eng., 119, 244–261, https://doi.org/10.1061/(ASCE)0733-9429(1993)119:2(244).
Peng, S.-H., and S.-C. Lu, 2013: FLO-2D simulation of mudflow caused by large landslide due to extremely heavy rainfall in southeastern Taiwan during Typhoon Morakot. J. Mt. Sci., 10, 207–218, https://doi.org/10.1007/s11629-013-2510-2.
Pradhan, B., 2010: Remote sensing and GIS based landslide hazard analysis and cross validation using multivariate logistic regression model on three test areas in Malaysia. Adv. Space Res., 45, 1244–1256, https://doi.org/10.1016/j.asr.2010.01.006.
Shamsudin, S., S. Dan’azumi, and A. Ab Rahman, 2011: Uncertainty analysis of HEC-HMS model parameters using Monte Carlo simulation. Int. J. Modell. Simul., 31, 279–286, https://doi.org/10.2316/Journal.205.2011.4.205-5487.
Shi, P., 2016: Hazard, disaster and risk. Disaster Risk Science (in Chinese), P. Shi and C. Jaeger, Eds., IHDP/Future Earth-Integrated Risk Governance Project Series, Beijing Normal University Publishing House Press, 1–48.
Totschnig, R., W. Sedlacek, and S. Fuchs, 2011: A quantitative vulnerability function for fluvial sediment transport. Nat. Hazards, 58, 681–703, https://doi.org/10.1007/s11069-010-9623-5.
UNDRR, 2013: Global assessment report on disaster risk reduction. United Nations, 288 pp., http://www.unisdr.org/we/inform/publications/33013.
Wang, J., Y. Yu, Q. Gong, S. Yuan, and J. Chen, 2021: Geospatial distribution characteristics and key influencing factors of mountain torrents in Guangdong Province, China (in Chinese). J. Univ. Chin. Acad. Sci., 38, 611–623.
Wenzhou Municipal Bureau of Statistics, 2021: Annual Survey Office of the National Bureau of Statistic in Wenzhou (in Chinese). Wenzhou Statistical Yearbook, China Statistics Press, 14 pp.
Wu, J., N. Li, and P. Shi, 2014: Benchmark wealth capital stock estimations across China’s 344 prefectures: 1978 to 2012. China Econ. Rev., 31, 288–302, https://doi.org/10.1016/j.chieco.2014.10.008.
Wu, J., G. Han, H. Zhou, and N. Li, 2018: Economic development and declining vulnerability to climate-related disasters in China. Environ. Res. Lett., 13, 034013, https://doi.org/10.1088/1748-9326/aaabd7.
Wu, X., L. Zhou, G. Gao, Z. Ji, and J. Guo, 2016: Flood depth-damage curves for urban properties considering disaster prevention and mitigation capabilities: Evidence from Lizhong Town, Lixiahe region, China (in Chinese). Prog. Geogr., 35, 223–231, https://doi.org/10.18306/dlkxjz.2016.02.009.
Xinhua News Agency, 2013: Typhoon Fitow caused heavy damage in Zhejiang Province, 18 counties and urban areas were flooded. Accessed 20 June 2020, http://www.gov.cn/jrzg/2013-10/09/content_2502813.htm.
Xu, L., Q. Li, Y. Wang, J. L. Huang, and Y. J. Xu, 2020: Analysis of the changes in debris flow hazard in the context of climate change (in Chinese). Adv. Climate Change Res., 4, 415–423, https://doi.org/10.12006/j.issn.1673-1719.2020.004.
Ye, M., J. Wu, C. Wang, and X. He, 2019: Historical and future changes in asset value and GDP in areas exposed to tropical cyclones in China. Wea. Climate Soc., 11, 307–319, https://doi.org/10.1175/WCAS-D-18-0053.1.
Yue, W., J. Gao, and X. Yang, 2014: Estimation of gross domestic product using multi-sensor remote sensing data: A case study in Zhejiang Province, East China. Remote Sens., 6, 7260–7275, https://doi.org/10.3390/rs6087260.
Zhang, J., Z. X. Guo, D. Wang, and H. Qian, 2016: The quantitative estimation of the vulnerability of brick and concrete wall impacted by an experimental boulder. Nat. Hazards Earth Syst. Sci., 16, 299–309, https://doi.org/10.5194/nhess-16-299-2016.
Zhang, J., W. Xu, X. Liao, S. Zong, and B. Liu, 2021: Global mortality risk assessment from river flooding under climate change. Environ. Res. Lett., 16, 064036, https://doi.org/10.1088/1748-9326/abff87.
Zhang, P., H. Ren, M. Zhang, W. Hu, L. Shi, and B. Zhang, 2006: A preliminary study on the zoning of flash flood disaster prevention and control in China (in Chinese). J. Water Resour. Res., 27, 15–18.
Zhang, Y., Y. Wang, Y. Chen, F. Liang, and H. Liu, 2019: Assessment of future flash flood inundations in coastal regions under climate change scenarios—A case study of Hadahe River basin in northeastern China. Sci. Total Environ., 693, 133550, https://doi.org/10.1016/j.scitotenv.2019.07.356.
Zhao, M., W. Cheng, C. Zhou, M. Li, N. Wang, and Q. Liu, 2017: GDP spatialization and economic differences in South China based on NPP-VIIRS nighttime light imagery. Remote Sens., 9, 673, https://doi.org/10.3390/rs9070673.
Zhao, Y., X. Zou, L. Cao, and X. Xu, 2014: Changes in precipitation extremes over the Pearl River Basin, southern China, during 1960–2012. Quat. Int., 333, 26–39, https://doi.org/10.1016/j.quaint.2014.03.060.
Zheng, H., Z. Gui, H. Wu, and A. Song, 2020: Developing non-negative spatial autoregressive models for better exploring relation between nighttime light images and land use types. Remote Sens., 12, 798, https://doi.org/10.3390/rs12050798.
Zhu, H., Z. Jiang, J. Li, W. Li, C. Sun, and L. Li, 2020: Does CMIP6 inspire more confidence in simulating climate extremes over China? Adv. Atmos. Sci., 37, 1119–1132, https://doi.org/10.1007/s00376-020-9289-1.
